PE1655MC network card problem

[pzero]

Hi all,
I have a few PowerEdge 1655MC Blade servers running Red Hat Enterprise 
Linux 3 ES U3.

Lately I am experiencing a problem with a NIC of one of them.

Basically, the server gets disconnected from the network in a random way 
and the network link goes up and down a lot of times, sometimes it comes 
back and it works fine for a few hours ...

Here are some entries from the '/var/log/messages' file:

---
Oct 15 13:38:00 myreplace kernel: tg3: eth0: Link is down.
Oct 15 13:38:40 myreplace kernel: tg3: eth0: Link is up at 1000 Mbps, 
full duplex.
Oct 15 13:38:40 myreplace kernel: tg3: eth0: Flow control is off for TX 
and off for RX.
Oct 15 13:39:09 myreplace kernel: tg3: eth0: Link is down.
Oct 15 13:46:32 myreplace kernel: tg3: eth0: Link is up at 1000 Mbps, 
full duplex.
Oct 15 13:46:32 myreplace kernel: tg3: eth0: Flow control is off for TX 
and off for RX.
---

The other Blade servers in the same chassis have no network problems at all.

I rebooted the server in question a few times and updated the PE1655MC 
integrated switch to the latest version with no luck.

Any idea on what's wrong?

Thanks.